The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes local TypeScript scripts included in the package to perform markdown chunking and content analysis.
The agent is instructed to run scripts/main.ts using the bun runtime or npx (as specified in SKILL.md).
These scripts are used to split large markdown files into smaller pieces for parallel processing by subagents.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data from external URLs and files for translation, creating a surface for indirect prompt injection.
Ingestion points: Untrusted content enters the agent context via URL fetching and file reading, as described in references/workflow-mechanics.md.
Boundary markers: Absent. The translation prompts defined in references/subagent-prompt-template.md do not include specific delimiters or instructions to ignore commands embedded within the source text.
Capability inventory: The skill possesses file-write capabilities (to the translate/ and output directories) and the ability to spawn subagents to process the ingested content.
Sanitization: No sanitization, escaping, or filtering of the external content is mentioned before it is passed to the translation prompts.

baoyu-translate